Suppose "abc" is a latent variable having 3 dimensions"x", "y", and "z". All the latent variables viz. "abc", "x", "y", and "z" have five items each. Now I want to measure relative contribution of "x", "y", and "z" to the overall score of "abc". what statistical method can be applied? Could I consider "abc" as dependable variable and 3 dimensions as independent variables and go for path analysis through SEM?
I would look at the differences between ABC and each of its components.
ABC-A
ABC-B
ABC-C
ABC-AB
ABC-AC
ABC-BC
The larger this difference the weaker the contribution. Thus I would order these differences and that will give me a sense of the relative contribution of each component.
You could also build models. Response = ABC + A, Response = ABC + BC, and so forth. Here you are looking at the significance level of the ABC term in the model when another factor is present.
A better option would be to use principal components, but there is no promise that the constructs that PCA gives will be the same as the ones that you want to use.
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Structural Equation Modeling